Various scientific and scholarly articles are referred to in parentheses throughout the specification. These articles are incorporated by reference herein to describe the state of the art to which this invention pertains.
Pathologists, who examine disease progression and analyze tissue samples for abnormalities, including cancer, have determined that a cellular condition called dysplasia, which refers to abnormal formation or maturation of cells, can potentially identify cells in a pre-cancerous condition. Unchecked, dysplasia can progress through mild, moderate and severe stages and eventually to cancer. About one in seven of the moderate cases of dysplasia will progress to cancer, and as many as 83% of cases with severe dysplasia have been reported to progress to cancer, depending on the types of cells involved. However, removal of mild and moderate dysplasias greatly reduces the development of cancer. In the lung, removal of dysplastic cells not only greatly reduces the formation of cancerous cells, but in some cases pulmonary tissue will return to a normal morphology.
In general, the earlier cancers are detected, the better the prognosis is for patient survival. If breast cancer is detected early when it is still localized to a single mass, the five-year survival rate is more than 96%. When it has spread to a distant location, the five-year survival rate is less than 20%. For lung cancer, when it is detected as a single mass the 5-year survival is more than 46%. When it has spread, the five-year survival is less than 14%. For cervical cancer, additional improvement in survival occurs when pre-cancerous changes are found and treated before developing into a more severe stage (Boring and Squires 1993, CA Cancer J Clin 43:7-26 and Ferguson 1990, Hematol Oncol Clin N Am 4:1053-1168).
Lung carcinoma is presently the leading cause of cancer mortality among men and women in the United States (Wingo et al. 1995, CA Clinical J Clin 45.8-30). In 1997, there were an estimated 160,000 deaths from lung cancer, accounting for 12% of all cancer deaths in U.S. men and 2% in U.S. women (Boring & Squires 1993, supra). Lung cancer is also one of the most lethal types of cancer, as reflected in a five-year survival rate of only 14%. The poor prognosis for lung cancer patients, relative to other types of human cancer, is due largely to the lack of effective early detection methods. At the time of clinical (symptomatic) presentation, over two thirds of all patients have regional nodule involvement or distant metastases, both of which are usually incurable. In studies of patients with localized (Stage 0 or 1) lung cancer, however, 5-year survival rates have ranged from 40% to 70% (Boring & Squires, 1993, supra; Ferguson, 1990, supra).
Historically, the only diagnostic tests used to detect lung cancer before symptoms occur have been sputum cytology and chest radiography. As a consequence, the efficacy of these tests as mass screening tools has been extensively evaluated in studies over the past several decades. Both tests detect presymptomatic, earlier-stage carcinoma, particularly carcinoma of squamous cells.
Improvements in screening methods have largely centered around improving the utility of sputum cytology through technologic advances in microscopy. Sputum cytology requires a visual examination of a cell sample during which cell size, shape, organization, and a ratio between the size of the cell's nucleus and cytoplasm is used to determine the cell's morphology. Because this assessment of cell morphology requires a visual inspection and classification, the technique requires a significant amount of expertise on behalf of the clinical observer. Various investigations have been conducted with results suggesting that computer-assisted, high resolution image analysis enables detection of subvisual changes in visually normal nuclei associated with several tissue types (Montag et al. 1991, Anal Quant Cytol Histol 13:159-167; Haroske et al. 1988, Arch Geschwulstforsch, 58:159-168; Hutchinson et al. 1992, Anal Quant Cytol Estol 4:330-334). Computer-assisted analysis of DNA distribution in cell samples provided 74% correct morphological classification of nuclei without human review of the material and without the need for visually abnormal nuclei being present when compared with standard cytological testing.
The morphologic assessment of cytological specimens has also improved due to advances in the understanding of lung tumor pathology. Much of this work has centered on the identification of “biomarkers.” Biomarkers refer to a wide range of progressive phenotypic and genetic abnormalities of the respiratory mucosa which may be used in determining the potential for bronchial epithelium to fully transform into a malignant tumor. Markers have been broadly classified as morphological changes, immuno/histochemical markers for differentially expressed proteins, markers for genomic instability, markers of epigenetic change (e.g., abnormal methylation), and gene mutations (Hirsh et al. 1997, Lung Cancer 17:163-174).
The expression levels of these markers are now being evaluated in dysplastic and neoplastic cyto/histological tissue samples collected from high risk populations. Among those specimens currently being targeted for exploratory marker analysis is sputum. Interest in sputum samples for biomarker research has been generated from the long-held belief that exfoliated cells recovered in sputum may be the earliest possible indication of an incipient carcinoma, since lung cancer most frequently develops in the bronchial epithelium. Through application of sophisticated molecular genetic techniques (e.g., PCR-based assays), studies are providing evidence that selected biomarkers can be detected in sputum (Mao et al. 1994, Cancer Res 54:1634-1637; Mao et al. 1994, Proc Natl Acad Sci USA 91:9871-9875; Sidransky 1995, J Natl Cancer Inst 87:1201-1202; Tockman et al. 1988, J Clin Oncol, 11: 1685-1693; Tockman et al. 1994, Chest, 106:385s-390s).
Commercially available cancer screening or detection services rely on tests based on cytomorphological diagnosis by trained clinicians who look at each sample and determine the extent and identity of abnormal cell types. This process is not only expensive and time-consuming, it also introduces human judgment and therefore error into the procedure. Recently, a method has been developed for detecting cancerous cells of the lung through use of 5,10,15,20-tetrakis (carboxyphenyl)-porphine (TCPP) (U.S. Pat. No. 5,162,231 to Cole et al.) This method relies on the propensity of cancerous cells to accumulate TCPP from their environment in a greater amount than non-cancerous cells. Upon incubation of a cell sample for 6-24 hours with 200 μg/ml TCPP, the TCPP entered cells and bound to the perinuclear membrane and mitochondria of neoplastic cells. TCPP fluoresces under ultraviolet light, and cancerous cells may thereby be diagnosed solely by the intensity of fluorescence, without reference to morphology. The extension of the use of this compound to identifying pre-cancerous tissue conditions (e.g., dysplastic cells) would permit screening in high risk populations to identify those individuals whose tissues are progressing toward invasive cancer conditions, and thereby facilitate catching the cancer or dysplasia at the most treatable stage. The desirable characteristics of such a screening method would be a procedure that is rapid, inexpensive, and requires a minimum of technical expertise.
For the foregoing reasons, there is a need for a technique and methodology for detecting dysplastic cells in their earliest stages. In addition, there is a need for a technique that can provide highly reliable diagnostic results and that does not rely on the subjective analysis of the clinician performing the diagnosis.